• Home
  • News
  • Background
  • From design, supply, distribution, marketing to retail: How AI is being used in the fashion industry

From design, supply, distribution, marketing to retail: How AI is being used in the fashion industry

By Esmee Blaazer

loading...

Scroll down to read more

Background

ZEGNA X developed on Microsoft press release 17 March 2023 Credits: Zegna Group

Attention fashion brands and retailers: how you can use Artificial Intelligence to make your business more intelligent and/or efficient

“Artificial intelligence (AI) can make the fashion industry smarter,” said Matthew Drinkwater, head of the Fashion Innovation Agency at the London College of Fashion and specialist in new technologies and their application in creative industries such as fashion, on stage at Web Summit, the tech conference that took place in Lisbon in November 2023.

“Using AI, fashion businesses – and therefore fast fashion giants – can offer consumers what they want, when they want it,” said Drinkwater. In other words, predicting and delivering products or services at the ideal moment. AI also has countless other applications. Consider, for example, assistance in the sketching and design phase. Or improving the retail experience in the physical store, through experience, interaction and (hyper)personalisation.

In this background story, FashionUnited highlights a number of examples.

Glossary:
  • Artificial intelligence (AI): This refers to computer systems that are able to perform tasks that normally require human intelligence, such as analysing data and making recommendations. AI systems learn and function based on the data they receive.
  • Generative AI: A technology that is able to produce images, text and audio based on instructions/commands – so-called “prompts” – from humans. It can therefore generate creative, new content. For example, it can write a new story, create a new song, create a unique work of art or a piece of clothing.
  • Predictive AI: This form of AI predicts future outcomes based on historical data, such as predicting the weather or estimating purchases. You can also use predictive AI to improve things, such as risk assessment and optimising production processes.

1. Predicting consumer demand

AI trend forecasting

AI is well-suited for mapping consumer preferences and making trend forecasts for the fashion industry. By using AI and data analysis, patterns, trends and emerging styles can be detected. Heuritech, an AI fashion trend forecasting company from Paris, scans millions of images on social media to gain insights. Heuritech's AI can distinguish more than 2,000 different ‘fashion details’ such as textures, prints and patterns, and 500 different colours when identifying trends. Well-known brands such as Louis Vuitton, Dior, Adidas, New Balance, Prada, Decathlon and Chalhoub work with the company.

A representative of Heuritech informed FashionUnited in writing that the company has just implemented a major update to its forecasting capabilities. “Thanks to two years of hard work, we can now predict trends up to 24 months in advance (previously it was up to 12 months in advance, ed.).”

The text continues below the screenshots of the platform that show this new feature.

Heuritech trend forecasting - Leopard next 24 months,Credits: property of Heuritech
Heuritech Trend Radar Credits: property of Heuritech

Retail intelligence

Retail intelligence platforms also use AI. Trendalytics (with clients such as Free People, StitchFix, Macy's and Saks Fifth Avenue), for example, uses Google Trends, social media and e-commerce market data to predict which trends are emerging or declining, the company wrote on its website. Edited, meanwhile, (with clients such as Puma, Bonprix, J.Crew and PVH) helps retailers achieve better results through AI-driven automation and insights in the field of market and business information.

Using AI for buying

WGSN launched an advanced AI-powered platform specifically for buyers earlier this month. The aim is to help them in different phases of the buying process: pre-planning, development and in-season, the well-known trend agency wrote in a press release.

Monisha Klar, director of fashion at WGSN, said: “Buyers have relied heavily on their own internal data until now. But because you don't have data on items you haven't sold, some buyers make decisions based on gut feeling, which may or may not be successful. I think WGSN is in the ideal position to give buyers the confidence and reassurance they need to support trends based on quantitative data (..).”

These background stories tell you more about the buying process of independent retailers and how they use figures in this process:

WGSN states that by switching to a demand and data-driven planning and buying model, overproduction can be reduced by 5 to 15 percent and waste in the sector can be reduced.

The text continues below the photo

WGSN Fashion Buying Platform Credits: WGSN (At the very bottom of this piece you will find a full size image of the WGSN Fashion Buying Platform, ed.)

2. Inventory management and distribution optimization with the help of AI

AI can also help optimise inventory: getting the right fashion items to the right place at the right time. This way, excess inventory, or so called remainders (unsold goods, which are usually destroyed, ed.) can be reduced and the risk of sold-out products can be minimised.

The Amsterdam-based tech company Wair specialises in making predictive AI forecasts for inventory and distribution optimisation. Clothing retailer Shoeby, active in The Netherlands and Germany, uses these services. With results, as was highlighted earlier this year in an episode about AI on the shop floor of Business Wise. “Better sell-through and the remainders has been reduced,” says Adrie Kappen, manager operations, to Business Wise AI expert Job van den Berg. “On all fronts, costs, return (profitability, ed.) and sustainability, it has brought us a lot.”

Dutch men's fashion clothing chains Van Dal Mannenmode and Only for Men (OFM) as well as sports retailer Daka are also clients of Wair.

The text continues below the image

Wair Credits: property of Wair

3. AI in fashion design and the product development phase

AI offers many possibilities for design and product development.

AI on the Catwalk

The New York label Collina Strada and British fashion designer Christopher Kane, for example, have already presented AI-generated prints on the catwalk. Ukrainian fashion brand Kseniaschnaider showed AI-made denim silhouettes during London Fashion Week last year.

The text continues below the video showing AI-generated animal prints from Christopher Kane):

Recently, on 17 July, the event "Culture x AI" took place at the V&A Museum in London, where clothing made with the AI-fashion design tool ‘AiDa’ was presented by five fashion designers from Hong Kong.

The text continues below the images

AI-generated fashion from Derek Chan, Wilson Choi, Tak Lee, Sophia XinLi and Aries Sin showed at the V&A Museum in July 2024. This specific look is a design from Sophie XinLi. Credits: Culture x AI 2024-2025: Culture and Future Mode .
AI-generated fashion from Derek Chan, Wilson Choi, Tak Lee, Sophia XinLi and Aries Sin showed at the V&A Museum in July 2024. This specific look is a design from Aries Sin. Credits: Culture x AI 2024-2025: Culture and Future Mode

AI fashion designer tools such as AiDa and Refabric

AiDa is an AI-based design assistant developed by AiDLab (short for Laboratory for Artificial Intelligence in Design, a collaboration between Hong Kong Polytechnic University and the Royal College of Art from the UK). The tool was launched with the Fashion X AI show in 2022 and made available to designers in Europe and Asia-Pacific. AiDa is currently in version 3.0.

AiDa supports designers by creating visual designs based on mood boards and sketches, which can make the design process 60 to 70 percent faster. The system can create original design collections based on images uploaded by the designer, such as mood boards. AiDA is able to distinguish 2,300 different colours.

Another AI-powered design tool specifically developed for the fashion industry is Refabric. The company is established by Seda Domaniç, former editor-in-chief of Vogue Turkey and the start-up has been selected for the LVMH acceleration programme, La Maison des Startups.

“Refabric can turn ideas into original designs by using prompts, inspirational images, sketches and patterns, which present variations of designs on different backgrounds and models in minutes,” the company said in a press release. Refabric already collaborates with fashion brands Aquazzura, Trendyol and DeFacto and the design school Istituto Marangoni Miami, a spokesperson informs us in writing.

Refabric also uses AI for 3D modelling to create digital prototypes of designs. This helps to reduce physical samples and production waste, thus contributing to efficiency and sustainability.

The text continues below the image and the video

Photo of AI Fashion Design platform Refabric Credits: property of Refabric

Using AI as a creative stimulus

AI should be seen primarily as an interactive ‘co-designer’, not as a replacement, as fashion innovator Calvin Wong, creator of AiDa stated. This is also the vision of Matthew Drinkwater of the London College of Fashion. He spoke about how AI was used in a 2019 MIT project to design dresses based on 40,000 images. “Instead of seeing AI as a replacement, the designers used it as a source of inspiration, which enriched their creative processes. Many of them still use AI to generate inspiration from behind their laptops,” said Drinkwater.

The text continues below the video

4. Improving retail experiences

“Unfortunately, high streets look the same all over the world,” said Drinkwater at the Web Summit. “Retail is about creating a sense of discovery—falling in love with a product, like a sweater, and building an emotional connection.” AI is a powerful tool to transform stores into spaces where the experience is central, rather than just the product, he said. “My vision is that stores in the future may not even have physical products ; it's all about providing an immersive experience,” said Drinkwater at the tech conference.

Also read: Stores without products: the future of retail?

Retail expert Tim Nash, head of IPOS Ignite and curator of Shop Drop Daily, also believes that AI can transform physical retail into personalised, interactive experiences. “Imagine walking into a store where the layout adapts to your preferences in real time, with tailored recommendations,” he told FashionUnited. “By integrating AI with the tangible, sensory experience of shopping, brands could create an environment that feels almost magical, where every visit feels like it was designed specifically for you.” It has the potential to completely change customer engagement and loyalty, said Nash.

A practical example is the Italian luxury menswear brand Zegna, which uses AI and data with Zegna X to offer a personalised shopping experience (see header photo). Customers receive advice from style consultants and can view different clothing combinations using a 360-degree configurator. This configurator can put together more than 49 billion outfits.

Personalisation or on demand production

AI can also be used in other ways for customisation. Consider, for example, product personalisation. Balmain recently used AI to personalise their Unicorn sneakers, in collaboration with Space Runners and artist Ant Kai, using generative AI to offer customers unique customisation options.

Another possibility is to use AI for on demand production. This means that garments are only produced after an order has been placed (where normally fashion items roll out of the factory in large quantities before it is known how much demand there will be from consumers). Desigual recently launched an on demand collection. The Spanish fashion brand has also used AI technology for optimising the production process by calculating exactly how much material is needed and by starting production as soon as there is demand. This reduces material waste and therefore further contributes to sustainability.

In theory, AI even has the potential to reduce or halt a major problem in the fashion industry, namely, overproduction.

The text continues below the photos

Balmain x Space Runners, Unicorn sneakers by Ant Kai. Credits: Space Runners.
Desigual uses artificial intelligence to design an On Demand Collection Credits: Desigual.

5. AI for e-commerce

There are countless applications of AI for online shops. By analysing customer buying behaviour, preferences and search queries, AI can, for example, suggest products that are most relevant to each individual user. Many online shops, from Asos to Amazon, use this to increase customer satisfaction and sales.

AI can also be used for virtual fitting, as already offered by Google, for example, and tools such as YourFit, Artisse or Vue AI make possible.

It can also help customers find the right size or fit. French fashion brands Vanessa Bruno, Sandro, Maje and Ba&sh collaborate with the AI-driven ‘sizing tool’ Fringuant, Balmain with Bods, Bershka with 3DLook and Marc Cain with Sizekick.

These AI solutions can contribute to improving the customer experience, increasing conversion (the percentage of visiting customers who make a purchase) and reducing returns.

The text continues below the images

Sizing tool Fringuant Credits: Fringuant
Sizing tool Fringuant Credits: Fringuant
Credits: 3DLook x Bershka

AI can also be used to make the presentation of products more inclusive and appealing to a wider audience. The Dutch start-up Lalaland, for example, specialises in creating virtual models with different body shapes and skin tones. Levi's experimented with this, but also received criticism.

Because using AI to generate diverse models was seen as a missed opportunity to hire people/models from different backgrounds and body shapes.

The text continues below the image

Credits: An AI-generated model. Image via Levi Strauss & Co from the FashionUnited archive (March 2023)

AI can also be used to create e-commerce content. The company Wair (mentioned earlier in relation to inventory and distribution, ed.) has developed an AI tool for creating e-commerce content. You can simply upload photos of new products, after which AI scans the images, adds detailed tags and metadata, and automatically generates product titles and descriptions based on this. “Shoe brand Steve Madden, Dutch large retailer Roetgerink and clothing chain Shoeby use our tool,” said a spokesperson for Wair.

The text continues on the image

Product describer AI tool E-commerce Content Creation from Wair Credits: property of Wair

6. Marketing

AI is also increasingly being used for marketing.

Advertising campaigns

Luxury brands such as Etro and Misela have used generative AI for advertising campaigns, i.e. commercials in which no real models or photoshoots have been used. Recently, fashion chain Mango also launched an AI face for its new teen line Mango Teen. It can save a lot of costs and work that comes with traditional shoots, but there are of course also ethical issues.

Read also:

Social media

AI tools are playing an increasingly important role in automatically generating and curating content for social media. With AI-powered text and image generation tools, brands can quickly create posts, videos, images and even entire campaigns that seamlessly align with their branding and business goals. This not only saves time and resources, but also ensures that the content remains consistent with the brand message.

Recently, the start-up Pre launched the tool "Pre Social" at the ‘International Conference on Social Media and Society’. This tool can predict with 80 percent accuracy how many likes a post on social media will get, as Vogue Business highlighted in a recent article. This allows brands and content creators to optimise their posts by determining in advance which content is likely to generate the most engagement and thus optimise their performance.

The text continues below the images

ADV AI campaign for SS24. Credits: Etro x Silvia Badalotti.
AI generated campaign Mango Teen Credits: Mango

In Conclusion

New technologies transforming the future of fashion

Drinkwater was asked on the Web Summit stage where AI will be in three to five years. His response highlighted how fast developments are going: "It's hard to look further than a few months ahead," he said in November 2023. "Three to five weeks ago I couldn't have predicted where we are now."

“What I do know is that AI will have a significant impact on the fashion industry and that we are at a tipping point,” Drinkwater continued. My advice: experiment with AI, test and learn! Ride the wave.”

More reads about AI:
WGSN Fashion Buying Platform full image Credits: Credits: WGSN (source 'WGSN to Launch Fashion Buying Platform' news, August 2024)

Sources:
- Web Summit talk ‘How fashion brands are using GenAI’ with Matthew Drinkwater, Head of the Fashion Innovation Agency at the London College of Fashion & Lauren Puricelli, co-founder and CEO of GenAI powered luxury tech studio Dress for Good on 14 November 2023.
- Written input from Heuritech spokesperson on 22 August 2024.
- WGSN press release ‘WGSN to Launch Fashion Buying Platform’ from 5 August 2024.
- Business Wise.nl ‘AI at work, what AI can do in the retail sector’ by Job van den Berg, 1 April 2024.
- Written input from WAIR spokesperson from 23 August 2024.
- AFP article ‘AI 'no substitute' for a fashion designer's creativity’ from 18 September 2023.
- Culture X AI press release ‘Press Release | AI Pioneers and Fashion Visionaries Illuminate the Runway’ from 19 July 2024.
- InnovationHub Hong Kong website, HK technologies, and specifically the page ‘AI-based Interactive Design Assistant for Fashion(AiDA), accessed in August 2024.
- Reuters article ‘In Hong Kong, designers try out new assistant: AI fashion maven AiDA’ from 28 December 2022.
- Refabric press kit and written input from the Refabric spokesperson on 22 August 2024. - Written input from retail expert Tim Nash on 21 August 2024.
- The FashionUnited archive and specifically articles by Don-Alvin Adegeest, Rachel Douglass, Vivian Hendriksz, Regina Henkel, Danielle Wightman-Stone, Marthe Stroom (linked in the article text).
- Financial Times article ‘Sometimes reality is not enough’: fashion brands turn to AI’ by Emily Cronin, from 13 August 2024.
- Vogue Business article ‘How AI is transforming social media behind the scenes’, by Maghan McDowell, from 20 August 2024.
- Linkedin page PRE, accessed on 24 August 2024.
- Parts of this article text were created using AI tools, and subsequently edited.

More background stories:
AI
AiDA
Artificial Intelligence
Background
Buying
DESIGN
Experience
Fashion Education
Heuritech
Innovation
inventory
Lalaland
refabric
Retail
Shoeby
Sustainability
Technologies
Trends
unsold goods
Wair
Workinfashion